From genes to vaccine: A breakthrough in the prevention of meningococcal group B disease

Abstract

Although safe and effective vaccines exist for meningococcal serogroups A, C, W-135 and Y, no vaccine is available for routine use against disease caused by serogroup B (MenB). Consequently, MenB is now the most common cause of invasive meningococcal disease in Canada. MenB causes more than 80% of invasive meningococcal disease in infants and can occur at any age. The mortality and morbidity rates related to this disease are very high. Vaccine development against MenB has been hampered by the fact that MenB polysaccharide is not immunogenic in humans. Although vaccines derived from the outer membrane vesicle have been effective in controlling MenB outbreaks, such vaccines protect against the outbreak strain only. A new vaccine development strategy, reverse vaccinology, has led to the identification of genes coding for surface-exposed proteins, which are able to induce bactericidal antibodies against a broad range of MenB strains. A new vaccine containing a combination of these proteins has been tested in different age groups, in several clinical trials. The data available provide hope that control of MenB through routine vaccination will soon be possible.

Neisseria meningitidis is an encapsulated bacterium that colonizes the nasopharynx and occasionally penetrates the mucosa to cause invasive disease. There are five serogroups that cause most of the invasive diseases globally: A, B, C, W-135 and Y. Invasive meningococcal disease (IMD) is an uncommon, yet serious illness that affects approximately 235 Canadians each year: the average annual incidence between 1995 and 2006 was 0.77/100,000 (1). However, the rate of serious sequelae and death, particularly in infants, is high (1–4). In spite of prompt diagnosis and treatment, 7.2% of survivors experience at least one major disabling complication, and 10% to 15% of patients die (4). Because of its severity, prevention by means of vaccination has been the long-sought goal.

Serogroup B (MenB) is now responsible for the majority of IMD in Canada as a result of the decrease in MenC disease following the introduction of routine immunization with group C conjugate vaccine from 2001 to 2005 (1). Between 1999 and 2006, the annual burden of MenB in Canada was approximately 115 cases per year, with infants younger than one year of age at highest risk. In 2006, the incidence of IMD caused by MenB for infants younger than one year of age was 6/100,000, representing 22 of the 113 MenB cases reported for that year; children one to four years of age represented the next highest risk group, with an incidence of 1.31/100,000 and 18 cases, and adolescents were third at 0.97/100,000 and 21 cases (1).

These rates underestimate the actual burden of MenB disease. Case confirmation methods in Canada currently depend primarily on culture and/or Gram staining. When cerebrospinal fluid and blood samples are tested using polymerase chain reaction, which is not widely available in Canada, the number of confirmed cases in which the serogroup of the infecting strain is identified increases by up to 30% to 50% (5,6). The present article provides a brief overview of meningococcal vaccines used in Canada, describes the issues that have hindered vaccine development for MenB and, finally, describes reverse vaccinology, which is the breakthrough approach that led to the development of one of the candidate vaccines against MenB disease.

THE EVOLUTION OF MENINGOCOCCAL VACCINE DEVELOPMENT

The polysaccharide capsule of N meningitidis is a well-established virulence factor that enables the bacteria to survive in human sera (7). Polysaccharide vaccines against both MenA and MenC have been effective in controlling outbreaks, but the meningococcal polysaccharide vaccines, with the exception of MenA, are not adequately immunogenic in infants (8). Conjugating the polysaccharides to a carrier protein overcame this deficiency and led to the monovalent group C and quadrivalent serogroup A-C-Y-W135 conjugate vaccines now in clinical use. The effectiveness of conjugated MenC vaccination in Quebec over a seven-year period was estimated to be 87% (9). In the United Kingdom and Canada, conjugated MenC vaccines were also shown to confer herd immunity by reducing carriage in the adolescent population (10–12). Unfortunately, this proven method of vaccine development could not be used with MenB because its polysaccharide capsule is not immunogenic in humans: it has the same antigenic structure as a sugar molecule on the surface of fetal neuroblasts; consequently, the immune system does not recognize the MenB polysaccharide as foreign (13).

THE FIRST MENB VACCINES

The search for a nonpolysaccharide MenB vaccine was then directed toward the surface proteins. During growth, meningococci continually release outer membrane vesicles (OMVs), also known as blebs. It is possible to isolate and purify the OMVs for use as vaccines. The immunodominant protein on the OMV is porin A (PorA), which has more than 600 antigenic variants (13,14). Unfortunately, these variants are not immunologically cross-reactive. Therefore, while MenB OMV vaccines were effective in controlling prolonged epidemics caused by single MenB strains in Cuba, New Zealand and Norway, they were not suitable for routine use against endemic MenB disease, which is caused by multiple MenB clones. The effectiveness of the OMV vaccine used in New Zealand, in a universal program that targeted infants, children and adolescents, was estimated to be 80% for children younger than five years of age and 70% overall (13).

NEW MENB VACCINES

The identification of suitable vaccine antigens that would provide broad coverage against a wide range of MenB strains has been difficult because many of the surface proteins are antigenically highly variable (15,16). A novel approach referred to as reverse vaccinology takes advantage of new techniques in molecular biology (14). The genome of MenB was sequenced and gene analysis permitted the identification of bacterial proteins that could be used as potential vaccine candidates. Previously unrecognized proteins were selected for inclusion in the vaccine based on exposure on the outer membrane, their ability to induce bactericidal antibodies and their presence in a higher proportion of disease-causing strains. The resulting vaccine, referred to as the four-component MenB (4CMenB) vaccine, contains the OMV (PorA) from the New Zealand vaccine along with three recombinant proteins identified by reverse vaccinology: factor H binding protein (fHbp), Neisserial adhesin A (NadA) and Neisseria heparin-binding antigen (NHBA) (14,17). Studies have shown that each of these proteins plays an important role in the pathogenesis of IMD and is recognized by the sera of convalescing survivors, suggesting that the proteins are expressed during natural infection (18). The hope is that this vaccine will provide protection against almost all strains of MenB, and that MenB will not eventually be able to evade vaccine-induced killing by altering its surface proteins through mutation.

IMMUNE CORRELATE OF PROTECTION

The antibody assay that correlates with protection against invasive meningococcal disease is the serum bactericidal assay (SBA). Bactericidal activity depends on the presence of specific antibodies against one or more bacterial antigens, and the presence of complement. Two forms of SBAs for meningococci exist: one uses human complement (hSBA) and the other uses rabbit complement (rSBA). The original antibody correlate of protection, as well as results from vaccine field trials, used hSBA instead of rSBA. There is now international scientific consensus that an SBA titre ≥4 (when human complement is used in the assay) is the serological correlate of protection against all serogroups of N meningitidis (15,19–23). This correlate was first established by Goldschneider et al (22), by showing that United States Army recruits who lacked detectable bactericidal antibodies on arrival for basic training had a 50% probability of developing invasive disease if they became carriers of the epidemic MenC strain. Recruits with hSBA titres ≥4 who acquired the epidemic MenC strain were protected against IMD. Goldschneider et al also showed that an inverse relationship existed between age-related incidence of disease and age-related prevalence of hSBA titres ≥4 against MenA, MenB and MenC. Subsequent studies with polysaccharide vaccines, conjugated vaccines against MenA and MenC, and OMV vaccines against epidemic strains of MenB have confirmed the correlation between the presence of complement-mediated SBA and vaccine efficacy against IMD (19–21).

Evaluation of the immunogenicity of the 4CMenB vaccine has been based on this proven correlate of protection, hSBA. The hSBA responses were measured for each antigen separately by selecting reference MenB strains for the assay that express only one of four vaccine proteins. Seroprotection in infants and toddlers was defined as an hSBA titre ≥5, and in adolescents and adults as an hSBA titre ≥4. The 4CMenB vaccination schedules for infants, toddlers, adolescents and adults are based on results from clinical trials involving more than 8000 subjects, and are described below.

4CMENB VACCINATION SCHEDULES, IMMUNOGENICITY AND ADVERSE EVENT PROFILE

For infants, primary series administered either at two, four and six, or at two, three and four months of age, were equally immunogenic. Over 99% of subjects achieved protective titres for fHbp and NadA, and 79% to 87% achieved protective titres against the OMV at one month after the completion of the primary series (24). Titres measured at 12 months waned considerably, indicating a need for a booster to maintain protection through the high-risk period. A fourth dose given to infants at 12 months produced a robust anamnestic response, with an 11- to 19-fold increase in hSBA titres (25). This was consistent with an earlier study involving infants, demonstrating the immunogenicity of 4CMenB in a schedule of two, four and six months, with a booster at 12 months (18).

In toddlers, a two-dose schedule of the 4CMenB vaccine, separated by two months, administered either concomitantly or sequentially with the measles, mumps, rubella and varicella (MMRV) vaccine, was highly immunogenic, with 96% to 100% achieving an hSBA titre ≥5, for fHbp, NadA and PorA, measured one month after the second dose (26). In adolescents, 4CMenB was highly immunogenic in a two-dose vaccine schedule, separated by one or two months, with over 99% of 1631 immunized adolescents attaining hSBA titres ≥4 for all antigens in the 4CMenB (27). In an open-label study of 54 adult laboratory workers, 89% reached hSBA titres ≥4 for each antigen after two doses of the 4CMenB vaccine (28).

An analysis of adverse events in infants revealed that fever (axillary temperature ≥38°C) was more frequent in subjects receiving concurrent 4CMenB and routine vaccines (44% to 61%) than in those given either the 4CMenB vaccine alone (26% to 41%) or routine vaccines alone (23% to 36%). The frequency of fever was similar after each dose of the 4CMenB vaccine. There were no significant differences in rates of systemic adverse events. Injection site redness and pain occurred in most recipients of the 4CMenB vaccine, but large local reactions (>100 mm) were rare (<1%) (29). In a separate study of 3630 infants (30), fever was evaluated daily for all subjects after each dose. Approximately 40% of infants experienced fever ≥38.5°C after doses 1 and 2 of the 4CMenB vaccine, and 20% after dose 3. Less than 1% had fever ≥40°C. Reports of fever at all doses were most common at 6 h postvaccination, ranging from a maximum of 40% after dose 1 to a minimum of approximately 25% after dose 3. By day 2, reports of fever were reduced significantly, ranging from a maximum of approximately 30% after dose 2 to a minimum of approximately 15% after dose 1. By day 3, fever was resolved in all groups.

Prophylactic acetaminophen given 0 h, 4 h to 6 h, and 8 h to 12 h following vaccination was effective in reducing the rate of fever ≥39.5°C from 8.2% to 3.3%, with no significant differences in immune responses to either the 4CMenB or the routine vaccines (DTaP-HBV-IPV/Hib and Pnc7) (31).

In adolescents who were given either one, two or three doses of the 4CMenB vaccine, adverse events were slightly elevated compared with placebo. The most common local reactions after any dose were pain (79% to 93%, versus placebo 49% to 86%) and erythema (48% to 56%, versus placebo 29% to 40%). The most common systemic reactions were malaise (44% to 58%, versus placebo 22% to 48%), myalgia (36% to 47%, versus placebo 18% to 41%) and headache (37% to 50%, versus placebo 21% to 37%). No severe adverse events were reported. Fever >38°C was reported at a rate similar to the placebo (1% to 5%) (27).

ESTIMATING VACCINE COVERAGE AGAINST DIVERSE MENB STRAINS

The rarity of MenB disease, and the rarity and unpredictability of MenB epidemics makes clinical trials with efficacy end points impractical and, possibly, unethical given the presence of an accepted immunological correlate of protection. Vaccine effectiveness will likely be evaluated post-licensure through active surveillance of MenB incidence following the introduction of routine immunization. Therefore, it will be important to have a method to assess the degree to which the protein antigens in the vaccines match the proteins on the surface of the MenB strains that cause invasive disease, both before licensure to estimate the potential benefit of vaccination and after licensure to detect any changes in the surface proteins that may enable the strains to evade killing by vaccine-induced antibodies.

To estimate the potential coverage of the 4CMenB vaccine, a new assay system has been developed, which is referred to as the Meningococcal Antigen Typing System (MATS) (23). This assay is an antigen-specific ELISA with a polymerase chain reaction component to identify the PorA variant. MATS was developed by Novartis, but it has been adopted by Public Health Laboratories in the United Kingdom to estimate a potential match of the circulating MenB strains to the vaccine antigens. This assay system has also been provided to the National Microbiology Laboratory in Canada.

Essentially, MATS will quantify and characterize the protein concentration of three of the vaccine antigens (fHbp, NadA and NHBA) on any MenB strain tested. For each of these vaccine antigens, there is a validated level of expression that has been correlated with killing in SBA. For PorA, simply the presence of the correct variant will predict killing because it is such a dominant protein on MenB. Once a MenB strain is tested, the MATS will give a result of ‘predicted to be killed or not’ by the vaccine. The MATS is able to evaluate many MenB strains at once, which enables the rapid estimation of concordance between the antigens in 4CMenB and those in diverse MenB strains isolated from patients. The results of MATS are not to be equated with clinical efficacy: they are, rather, an estimate of the proportion of circulating MenB strains, which could be expected to be killed by antibodies induced by the MenB vaccine.

To validate MATS correlation to hSBA and its reliability to predict strain killing, a large panel of disease-causing MenB strains, isolated in Western Europe and North America, were tested using MATS and hSBA (23). The hSBA was performed using sera from the infants (who received three or four doses), adolescents and adults who participated in the clinical trials. Not only did the study validate the use of MATS to provide an accurate, quick estimate of strain killing, the study also showed that in adults, the probability of killing found in the SBA was correlated to the number of vaccine antigens present on a strain. MenB strains positive for one of the four vaccine antigens had a probability of being killed of approximately 85%, while strains expressing two or more proteins had a 96% probability of being killed. Interestingly, strains that were not positive for any of the vaccine antigens still had a 67% chance of being killed in SBA, implying that killing may result from the combined action of low titres of bactericidal antibody directed against several antigens. This indicates that MATS will provide a conservative measure of vaccine coverage of MenB strains.

While the panel of strains in the present study did not represent the disease-causing strains of any particular country, MATS estimated that the vaccine antigens in 4CMenB were matched to approximately 75% of isolates that caused invasive disease in 13-month-old infants and approximately 83% of isolates in adults (23). MenB isolates responsible for causing invasive disease in Canada are currently being tested, using MATS, to provide an estimate of the potential coverage of 4CMenB in Canada.

RESEARCH ISSUES

Outstanding issues include the following:

• Effectiveness of the vaccine in preventing invasive MenB disease

• Immunogenicity of a reduced infant schedule of doses in the primary series

• Necessity and timing of booster doses in childhood and adolescence

• Effect of immunization of adolescents on MenB carriage, because preventing carriage in this age group may lead to herd immunity

• Effect of antibodies induced by the vaccine on other serogroups, all of which have the same outer membrane proteins as MenB (32)

• Acceptability to parents of yet another vaccine in infants

• Emergence of MenB strains with surface proteins that do not match the vaccine components